The world is changing. And so is the world of recruitment.

For too long recruiters and hiring managers were dependent on job boards and LinkedIn Recruiter. As a matter of fact, the recruitment industry hasn’t changed since 20 years.

Today, many recruiters and hiring managers are tested on their skills to work with profile databases; using the right filters, build boolean strings and basically turn a list of requirements into a mathematical equation of search parameters and confusion outcomes. Not what recruiters set out to do when they started their career.

Data should be crunched by machines. The human in the loop is there for genuine human connection and to have meaningful empathetic conversations after the data has been crunched.

It’s weird when you think about it:

A hiring manager knows who they want to hire in their team.

They try to communicate to their recruiters who they’re looking for through a job description.

The recruiter then tries to interpret that job description but never has a complete understanding of who’s the right person for the job because they’re not the domain experts.

The recruiter tries to turn the job description into a set of search filters and starts screening profiles.

The search engine, like LinkedIn Recruiter, returns a set of profiles based on let’s say job title, keywords and location.

Those initial candidates in the search results are typically not the right matches because the search engine misses contextual information and understanding of crucial things like job history interpretation.

What’s wrong with this way of searching is the following:

• Relies on recruiter's knowledge about role, which is often very limited (you can’t expect a recruiter to deeply understand all relevant professional domains)
• Is in the end a math question: requirements factor x available talent (how many recruiters really understand how to manipulate the algorithm for optimal results?)
• Lacks contextual interpretation, signals, patterns and cross data field understanding (filters and keywords are used isolated from each other)
• Search is technical and databases have many limits which recruiters are not aware of (like when a recruiter uses a required keyword, not knowing this will limit results to profiles with that exact keyword)
• Errors by user interaction which compound (like typos and wrong use of search functions)
• Highly sensitive to bias of all sorts (recruiter inputs what they know, not what the hiring team needs)
• Time intensive and inefficient (using filters, iterating on search, reaching out, data management)

Not good! But now we identified the issue and ventilated our frustration with the status quo, let’s look at what the world of recruitment will look like this year and beyond.

Spoiler alert: 2024 is automated, scalable and powered by AI and specifically Large Language Models (LLMs) across the entire sourcing cycle.

The new recruitment: AI Recruiter

We’re excited to have launched the world’s first AI Recruiter, a recruitment agent who can find, screen and reach out to candidates independently and completely autonomously.

We call her Uwi.

Uwi:

• Can access 1 billion profiles on the web
• Then screens those profiles and tells you why they're a good fit for your job
• And finally reaches out to the top candidates on autopilot

You can try Uwi here.

Uwi is the first AI Recruiter in the world. So far there have been profile databases, filter systems and message sequence tools.

But no system has been focussed on completely automating the recruitment or sourcing cycle yet, they all require a lot of manual work and onboarding.

Uwi’s promise is that she can recruit people for you without you having to fight buttons, filters and user interface.

How recruitment on autopilot by Uwi works

Next to having access to public profile and email data, Uwi is deeply integrated with LLMs like the latest GPT and other models.

With this, all parts of the recruitment process that used to be very manual are now powered by the entire knowledge of the web, good at reasoning (very useful in screening) and able to execute on tasks autonomously.  

The most important pieces of the puzzle are:

1. Search
2. Screening
3. Outreach (or better; ‘Engagement’)

1. Search

The problem with the current way of doing a search for profiles is that it’s isolated (different search fields function separately from each other) and static (it’s a one time search without iteration). Uwi makes the search for matches contextual (a broader understanding of the context of the job and requirements) and more extensive (iterative searches done across platforms).

Instead of using filters and booleans in a database, you now just tell Uwi who you’re looking for, that’s it.

With this conversation, search becomes a discovery. Uwi asks the right questions based on your input which helps to get a more accurate and complete picture of what she will be looking for.

This completed view of the ideal candidate is crucial because often in job descriptions and requirements essential information is missed because it takes time to complete the information and simply hiring managers and recruiters just forget about important points to include in their reasoning and job description.

The recently launched AI Recruiter Uwi.

2. Screening

For screening, historically there haven’t been many solutions in the outbound recruiting process. Screening has been predominantly a manual process of looking at a profile in your search results and checking if they are indeed a good match for the job.

As recruiters know, this is a very time consuming process and not the most fun since it can have a very repetitive nature. Also, a lot of error and bias are introduced by screening manually because important information is missed or simply incorrectly interpreted.

Let’s say you’re looking for a startup employee. In your own manual screening you’re looking for signs of the right startup experience and mindset required to thrive in your company and open role. You’re screening profiles for known startups, keywords indicating startup experience and maybe other arbitrary criteria. You’re very likely to miss relevant information simply because you don’t know a startup company or you miss other factors indicating startup experience and interest.

This is just one of many examples, and every job description has this crucial context in it that is missed with traditional systems.

Uwi screens completely autonomously while accessing the world’s knowledge (through language models) to interpret the data on the profile and look for pros and cons with its broad understanding of language and context.

3. Engagement

Engagement is very much about sending the right message, to the right person, at the right time. What’s hard when doing this in a manual way is balancing the other tasks you have with reminding yourself and taking the time to send your messages.

Automating outreach just makes a lot of sense. And now we can get to a whole new level of personalization which is automatically generated in messages, we don’t have to trade off message quality and personalization anymore with scalability, we can do both at the same time by automating the entire process of outreach and follow ups.

You can’t send a highly personalized automatically generated message if you don’t link the engagement process to the search and screening process. You need the data and understanding of the job and the matched profile to have the AI generate a relevant personalized message.

Uwi engages talent on autopilot with hyper personalized generated messages that are completely matched to all profile and job data.

Now is the time to power your own process by AI. Try Uwi now.